X-ray luminescence computed tomography (XLCT) is a novel and promising molecular imaging modality. In XLCT, collimated X-ray photons excite nanophosphors that can be functionalized to induce luminescence light photons that are measured for tomographic imaging. Thus, XLCT promises to integrate spatial resolution of X- ray micro-CT and molecular sensitivity of optical imaging. However, this potential has not been implemented yet due to the following two challenges. First, it is rather difficult to collimate divergent X-rays into a thin pencil beam, and even with such a narrow beam the data acquisition process would take long time. Second, it remains an open question how to develop bright, safe and biologically relevant X-ray excitable nanophosphors. In this project, we will prototype a focused X-ray luminescence tomography (FXLT) system for spatial resolution of 150m, molecular sensitivity around 5M, and penetration depth sufficient for small animal imaging. The imaging time per transverse section is <2 minute with the radiation dose in the range of a typical micro-CT scan. Our pilot results demonstrate that we can use a polycapillary lens to focus X-rays from a focal spot of 55m into a dual-cone-shaped pencil beam of 78 m in the focused region, much thinner and much more intense than a collimated X-ray beam in the XLCT experiments reported in the literature. Hence, the spatial resolution of FXLT can be at least improved to ~150m. We will use 8 photomultiplier tubes (PMTs) to measure emitted optical photons on the mouse body surface at two emission wavelengths simultaneously. The parallel use of these single-photon-counting PMTs will substantially reduce the measurement time and the radiation dose. We will mount an X-ray tube with its lens on a linear stage which is in turn on a rotary gantry. On the same gantry, we will mount another X-ray tube and an X-ray photon detector for micro-CT for hybrid X- ray and optical imaging. We will develop compressed sensing algorithms for image reconstruction from micro- CT and FXLT data, aided by optimized combination of sparsity and correlation priors such as by minimizing dimensionality of the patch manifold of an image. The micro-CT images will guide the selection of regions of interest and allow attenuation correctness for quantitative FXLT. To enable and demonstrate the preclinical feasibility and merits of FXLT, we will synthesize bright nanophosphors with multiple emission wavelengths and surface functionalization. We will characterize and optimize these nanophosphors in terms of their emission efficiency, emission wavelengths, toxicity, and specificity. Finally, we will perform live mouse studies using our FXLT system, with an emphasis on longitudinal imaging of the EGFR density in deep tumor. Upon the completion of this project, we will have optimized and characterized the first-of-its-kind hybrid molecular imaging system FXLT. Also, we will have demonstrated the advantages of FXLT for preclinical molecular imaging. FXLT couples X-ray focusing and optical labeling for micro-CT resolution and optical sensitivity, and will be a vital molecular imaging tool for precision medicine.